• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

NDK NZ2520SDA oscillators group buy

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
not sure what it's worth, but here is some graph showing different measurements than the specs for SD version

The New Hypex Fusion Plate amps

measures about 10dB lower than on datasheet at 1khz.

What "bavmike" has done with his Purestream DAC is what I've done with my Wave I/O – I implemented LT3045’s and SDA’s… the LT3045's have, in fact, a very low output impedance.

What he’s doing with Panasonic ECHU PPS, I achieve with SILMIC Alpha Super Gold 2.2uF / 50V – incredible cap and of a right capacitance for digital audio power supply decoupling. Very expensive.

I was also going to ask Jocko for price on a couple of selected SDA’s. I like the plots that specify phase noise figures at 0.1 and 1Hz, but his equipment goes down to 0.01Hz which is incredible. At the other place, he wrote a fair bit on oscillators, crystals, low phase noise, influence of air disturbance / heat on oscillators & sound reproduction. A must-read, really!

The SDA’s I have in Wave I/O were randomly picked by Lucian as a next on a roll, but the board does sound incredibly enjoyable. My wife had time to sit and listen to a few well known tracks, after all the fine tuning I did in the last few days… – she was stunned at the difference between the “old” SD Wave I/O, and this latest SDA version.

I’ll keep an eye on that Purestream DAC… hopefully more details will be revealed (or they have been already?)
 
Clearly, if plot from NDK are accurate, and you can hear a major improvement, then high phase offset noise matters.

i have to read more about audio socillators...it's tough for me to know the important parameters and analyse data...i will try to find jocko's work

Regarding the elna super gold....i was not aware of their existence. I am looking for some time for an upgrade to the nichicon KZ that i use as output coupling caps for my headamp....they may be interesting, though i like how neutral are the nichicon. I Would use film caps but hundreds of uF is just physically to big.....
 
then high phase offset noise matters.
May be... More important might be the difference in the 1 - 1000 Hz region.

Let me explain.

We did some experiments with various 10 MHz OCXO connected to a MC-3+.
The soundwise best two OCXOs had crossing phase noise. The better OCXO had lower phase noise in the region mentioned above and 10-20 dB more phase noise in the high frequency region.

For audio a clock is important, which is exact for the duration of a sound event. cf. eg. Relationship between audio equipment and crystal oscillator Sound quality and clock phase noise
image_04.png


With audio the word clock error in time is more intuitive to me. Therefor I like phase noise converted to time jitter. cf. eg. Converting Oscillator Phase Noise to Time Jitter.

RMS Phase Jitter for a given bandwidth describes how the time error is distributed. Using eg. Phase Noise to Jitter Calculator you can see how low the contribution of phase noise at higher frequencies is.
Example: ABLJO-100.00MHz:
10 Hz - 1 kHz: 2,639 ps
10 Hz - 1 MHz: 2,640 ps​

Ulli
 
Last edited:
Soldering Tips Wanted

Hi,

I just received the NDK clocks and they are very tiny. I plan to solder them onto the clock adapter boards by Ian. I understand that the NDK clock can be damaged by too much heat so I wonder what's the best way to solder them onto the adapter board w/o frying them.

I have temp-controlled soldering iron and hot air rework heat gun. At first I planned to use the hot air gun but then worried that it may fry the clock as I don't have a pre-heat platform.

Another idea is to tin the pads on the adapter pcb, place the clock on top, and then heat the pads one by one to hopefully reflow them. I am not sure if this will work though as the clock chip will be cold and is not heated directly.

Suggestions are most welcome!
 
If you tin all of the pads and warm them up one at a time you'll have trouble getting the oscillator to sit flat on the pcb.

What I do in such cases is tin one of the pads, than place the oscillator on top and melt the solder by heating the exposed part of the pad. Then I proceed to solder the rest of the pads by applying heat and solder to the exposed part of the pads. Assuming you've used a proper amount of soldering paste (which you should always do anyway) the solder will just flow to the underside of the oscillator.

But then again you can just tin all of the pads, place the oscillator on top and use the heat gun to melt the solder. I've also done it that way, with oscillators that I didn't really care about damaging. Never had a problem yet.
 
Ben Mah: I read your post. I actually ordered a hot air rework station but is having second thought as I don't have experience on it and practicing hot air rework using the NDK chips could be an expensive exercise. :cool:

Extreme_Boky: Lucian's soldering work is indeed impeccable. I'll give a go as I found soldering SMD components is becoming unavoidable in this hobby.
 
Hello Jack! Thank you for confirming the delivery of your SDAs! If it does help you, I'm using both soldering iron and hot air station to solder those tiny parts!
The temp on both is set at 330 degrees centigrade but frankly I have no clear idea if the actual temp is matching the one displayed on the gears. Either way, I'm usually cleaning the place where the oscillators will be soldered, in case there was an older solder there, especially the ones mostly based only on tin.
I'm mostly using the Sn63/Pb37 solder wire due to low melting point and solder wick to clean up the old solder. After that I'm cleaning the place with isopropyl alcohol and put some new flux in there. After that I'm using the soldering iron with a small tip to make four "new" solder bumps on the places where the new part will be soldered. The idea is to keep the bumps equal so that your oscillators be soldered as parallel with the PCB as possible.
Do not make those bumps too high because you can end up with shorts! Not to mention that the metallic package of the NDK oscillators will love to take a 'taste' of that melted solder!
After that, the NDK oscillator will be kept in place with tweezers and use hot air gun to heat up the place. Remember to NOT keep the air nozzle straight above the part! You have to heat up the surrounding copper pads as well by moving the nozzle in circular form. When the bumps under the SDA oscillators are melt then your part will get self-centered above the footprint.
Basically you can move the part with hot air flux. Speaking of air flux, my station is set around a scale of 2.5 from 6 (maximum).

As you can see I'm not using solder paste of any kind because in my experience, the solder paste does not make a good bundle with hot air station! there are some exceptions though but most of the good results are involving preheaters as well... but, in my opinion, you don't need those for your work!
I hope it helps!

EDIT: I forgot to say that the nozzle tip I'm using is rather small, I guess it has the diameter of the NDK SDA itself. This way I can work on small areas without affecting the surrounding parts too much (by temperature). If the nozzle you're using does have a bigger diameter then you can compensate by rising the value of the air flux and then by rising the temp.
 
Last edited:
Lorien: Thanks for the tips.

I am happy to report that soldering the SDAs are successful!

I took a different approach as my hot air rework station is still in transit so I tried using my Hakko hot tweezers to solder the SDAs. I tined all 4 pads with solder, placed the SDA on top, and then use the hot tweezers to clamp the SDA to place it in-place while heating up the pads at the same time.

The SDAs survived despite being heated by the hot tweezers at 350C directly for about 5 seconds. While I am at it I also soldered two Crystek 957 - super easy by comparison.

I verified the oscillators by connecting them to a lab power supply at 3.3V and then used a scope to check the output (see attached).

One question: The Ian adapters have three sets of pads on the underside for decoupling caps. I installed just one cap of 100nF supplied with the adapter. Is that sufficient?
 

Attachments

  • IMG_2240_Small.jpg
    IMG_2240_Small.jpg
    230.2 KB · Views: 619
  • IMG_2244_Small.jpg
    IMG_2244_Small.jpg
    271.3 KB · Views: 604
I decided to mix things together to be easier for anyone to read and update their preferences!

NZ2520SDA Group buy
(for 22.5792 MHz, 24.576 MHz, 45.1584 MHz and 49.152 MHz oscillators):
1. M_Baou - 1 pair of 4x
2. sontero - 1 pair of 4x
3. kvl - 2 pairs of 4x
4. wlowes - 1 pair of 4x

You can express your request whatever you want/like!
Kind regards,
L

Edit: @ kvl: got it, thank you!
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.